We will support projects 4, 5 & 6 by providing methodology development and microbial informatics services. We will: 1) Improve analysis methods for high-throughput gene-targeted metagenomics. Because shotgun metagenome analysis of complex microbial communities is still difficult and costly, we will extend our Functional Gene Pipeline (FunGene) to support additional functional genes of interest to the projects. Also we will develop improved software tools to efficiently process the higher output of new sequencing technologies now on the horizon, such as PacBio and lon Torrent. 2) Support RNA-Seq analysis of biodegrading strains and microbial communities. Best practices for analysis of RNA-Seq data are still being developed, and the application of RNA-Seq to microbial communities, meta-transcriptiomics, is even less developed. We will support both types of analysis using a combination of available tools and new tools we will develop. 3) Support metagenome analysis and develop methods for targeted metagenome analysis. By simplifying complex microbial communities, either through culture enrichment, targeted gene enrichment, or cosmid selection procedures, the cost and complexity of metagenomics can be greatly reduced. We will assemble a combination of available tools and novel methods and apply these to analyze metagenomes developed from biodegrading enrichment cultures and mouse gut enriched for genes and organisms of interest. 4) Develop functional gene metagenome and metatranscriptome enrichment methods. We will develop a method to simultaneously pre-enrich bacterial metagenomic DNA for large numbers of genes of potential importance in biodegradation and toxin response as a novel application of the massively parallel biotinolyated RNA oligonucleotide 'bait' and capture' technique pioneered by Gnirke et al. We will develop the probe sequences for a large number of genes of interest by extending our FunGene database. For highly diverse genes such as aryl dioxygenase genes, we will supplement this with sequences obtained via genetargeted sequencing. We will first validate the method on DNA target fragment length and percent identity to bait sequences. Then we will evaluate the methodology for quantitative metatranscriptome analysis.